A method for processing root system images

Abstract

One or more embodiments of this specification provide a root system image processing method, including obtaining a root system image, processing the root system image using a root system image processing model to obtain a root system feature image, and extracting root system phenotypic parameters based on the root system feature image. Using the method of this embodiment, the root system can be identified quickly, efficiently, and accurately, and various root system phenotypic parameters can be extracted, greatly improving the efficiency of scientific research work and providing a data basis for root system research.

Description

Technical Field

[0001] One or more embodiments of this specification relate to the field of image processing technology, and in particular, to a method for processing root system images. Background Art

[0002] Root system observation is an important basis for plant root system research. The minirhizotron imaging technology can be used to obtain root system images, but important root system phenotype parameters such as root length and root tip number still need to be identified by researchers with the naked eye, resulting in low efficiency and low accuracy. Therefore, how to automatically extract root system phenotype parameters using image processing technology is a problem to be solved in this field. Summary of the Invention

[0003] In view of this, the purpose of one or more embodiments of this specification is to propose a method for processing root system images that can automatically extract root system phenotype parameters.

[0004] Based on the above purpose, one or more embodiments of this specification provide a method for processing root system images, including:

[0005] Obtain a root system image;

[0006] Process the root system image using a root system image processing model to obtain a root system feature image;

[0007] Based on the root system feature image, extract root system phenotype parameters.

[0008] Optionally, the root system feature image is a root system probability distribution map;

[0009] Based on the root system feature image, extracting root system phenotype parameters includes:

[0010] Perform binarization processing on the root system probability distribution map to obtain a root system binary image;

[0011] Based on the root system binary image, count the number of root system pixel points;

[0012] Calculate the projected area according to the number of root system pixel points.

[0013] Optionally, based on the root system feature image, extracting root system phenotype parameters further includes:

[0014] Perform skeleton extraction processing on the root system binary image to obtain a root system skeleton image;

[0015] Based on the root system skeleton image, extract root length, root diameter, and root tip number.

[0016] Optionally, based on the root system skeleton image, extracting root length includes:

[0017] Determine the root centerline according to the root system skeleton image;

[0018] Count the number of pixel points on the root centerline as the total root length.

[0019] Optionally, after determining the root centerline, it further includes:

[0020] Determine the position of the root centerline in the root system image. Taking each pixel point on the root centerline as the center of a circle, and using the distance from the center to the boundary point between the root system and the soil as the radius to draw a circle, and taking the diameter of the circle as the root diameter corresponding to the root position of the pixel point.

[0021] Optionally, after determining the root diameter, it further includes:

[0022] Divide the root diameter grades according to the root diameters of different pixel points on the root centerline.

[0023] Optionally, after determining the root centerline, it further includes;

[0024] Count the number of end points on the root centerline;

[0025] Determine the number of root tips according to the number of end points.

[0026] Optionally, after extracting the root length, root diameter, and number of root tips, it further includes:

[0027] Calculate the root volume and root surface area according to the root length and root diameter.

[0028] Optionally, after extracting the root system phenotypic parameters, it further includes:

[0029] Calibrate the root system phenotypic parameters according to the preset resolution correspondence.

[0030] Optionally, after obtaining the root system binary image, it further includes:

[0031] Filter out interfering pixel points according to the preset filtering conditions.

[0032] As can be seen from the above, the root system image processing method provided by one or more embodiments of this specification obtains the root system image, uses the root system image processing model to process the root system image to obtain the root system feature image, and based on the root system feature image, extracts the root system phenotypic parameters, which can quickly, efficiently, and accurately identify the root system and extract various root system phenotypic parameters, greatly improving the efficiency of scientific research work and providing a data basis for root system research. Description of the Drawings

[0033] To more clearly illustrate the technical solutions in one or more embodiments of this specification or the prior art, the following will briefly introduce the accompanying drawings required for the description of the embodiments or the prior art. Obviously, the accompanying drawings in the following description are only one or more embodiments of this specification. For those of ordinary skill in the art, without creative efforts, other drawings can also be obtained based on these drawings.

[0034] Figure 1 Schematic diagram of the method flow for one or more embodiments of this specification;

[0035] Figure 2 Binary image of the root system for one or more embodiments of this specification;

[0036] Figure 3 Skeleton image of the root system for one or more embodiments of this specification;

[0037] Figures 4A - 4D Experimental result diagrams of root system phenotype parameter calibration for one or more embodiments of this specification respectively;

[0038] Figure 5A Comparison diagram of the root diameter grading identified by the naked eye and the automatically identified root diameter grading for one or more embodiments of this specification;

[0039] Figure 5B Root diameter grading diagram for one or more embodiments of this specification;

[0040] Figures 6A - 6C Schematic diagrams of the changes in root length, projected area, and average diameter of maize root systems for one or more embodiments of this specification respectively;

[0041] Figures 7A - 7D Comparison schematic diagrams of root length, projected area, average diameter, and number of root tips extracted at different resolutions for one or more embodiments of this specification respectively;

[0042] Figure 8 Schematic diagram of the device structure for one or more embodiments of this specification;

[0043] Figure 9 Schematic diagram of the electronic device structure for one or more embodiments of this specification. Detailed implementation manners

[0044] To make the purpose, technical solutions, and advantages of this disclosure clearer and more understandable, the following further elaborates on this disclosure in detail in combination with specific embodiments and with reference to the accompanying drawings.

[0045] It should be noted that, unless otherwise defined, the technical terms or scientific terms used in one or more embodiments of this specification should have the ordinary meaning understood by those of ordinary skill in the art to which this disclosure pertains. The "first", "second" and similar terms used in one or more embodiments of this specification do not denote any order, quantity or importance, but are only used to distinguish different components. Words such as "including" or "comprising" mean that the elements or objects appearing before this word cover the elements or objects listed after this word and their equivalents, without excluding other elements or objects. Words such as "connected" or "linked" are not limited to physical or mechanical connections, but may include electrical connections, whether direct or indirect. "Up", "down", "left", "right", etc. are only used to represent relative positional relationships, and when the absolute position of the object being described changes, the relative positional relationship may also change accordingly.

[0046] As Figure 1 shown, an embodiment of this specification provides a root system image processing method, including:

[0047] S101: Obtain a root system image;

[0048] In this embodiment, first, the micro rhizotron imaging technology can be used to obtain the root system image, and an in-situ root system scanning device is used to perform in-situ scanning imaging on the plant root system without damage, continuously monitoring and obtaining the root system images during the plant growth process.

[0049] To improve the accuracy of feature extraction, the obtained root system image is an image that only includes soil and the root system, and there are no obvious large cracks in the soil, and there are no interference factors such as obvious scratches, water mist, blurring, etc. on the scanning device that affect the image effect.

[0050] S102: Process the root system image using a root system image processing model to obtain a root system feature image;

[0051] In this embodiment, the obtained root system image is input into a pre-trained root system image processing model, and after being processed by the root system image processing model, a root system feature image is output. Among them, the root system feature image is a root system probability distribution map, and each pixel value of the root system probability distribution map represents the probability of being predicted as the root system. For example, the pixel value ranges from 0 to 255, the larger the pixel value, the greater the probability of being predicted as the root system, and the smaller the pixel value, the smaller the probability of being predicted as the root system.

[0052] S103: Extract root system phenotypic parameters based on the root system feature image.

[0053] In this embodiment, after obtaining the root system feature image, the root system phenotypic parameters are extracted through image processing, completing the automatic recognition of the root system and the automatic extraction of the root system phenotypic parameters.

[0054] The root system image processing method provided in this embodiment includes obtaining a root system image, processing the root system image using a root system image processing model to obtain a root system feature image, and extracting root system phenotypic parameters based on the root system feature image. Using the method of this embodiment, after obtaining the root system image, machine learning technology and image processing technology are used to automatically identify the root system and extract the root system phenotypic parameters, greatly improving the work efficiency and recognition accuracy.

[0055] Optionally, the root system image processing model is trained based on the U-net neural network model. The root system images of specific root systems are used as training data and input into the U-net neural network model for training to output root system feature images. The model divides the input root system image into multiple sub-images, adds noise, and determines whether each pixel point is a root system. The sub-images after segmentation enter the encoder-decoder for processing. After multiple fusion processes, multiple sub-images with enhanced root system features are obtained. Finally, the multiple sub-images are stitched together to form a root system feature image with the same size as the root system image. The output root system feature image has prominent root system features, and there are not significant differences in color differences, brightness changes, etc., providing high-quality images for subsequent image processing.

[0056] In some ways, during the training process of the model, cross-entropy is used as the loss function:

[0057]

[0058] where ak(x) is the score of pixel point x corresponding to feature channel k; K is the number of classes. In this example, the number of classes is 2, one class is the root system, and one class is the background outside the root system; P k (x) is the classification result of pixel point x, that is, pixel point x is the root system or the background.

[0059] Use a weighted penalty function:

[0060]

[0061] where l is the true label of pixel point x, used to mask the root system in the image; ω is the weight of pixel point x. In order to compensate for the frequency differences of different pixel points in a certain category, the weight function is pre-computed using GT examples, enabling machine learning to distinguish smaller boundaries.

[0062] The weight function is:

[0063]

[0064] where ω c: Ω → R is a weight map used to balance a certain type of frequency; d1: Ω → R represents the distance from a certain pixel point in the background to the nearest target boundary (the outer contour of the occluded root system area); d2: Ω → R is the distance from a certain pixel point in the background to the second-nearest target boundary. Set ω0 = 10 and σ ≈ 5 pixels. Additionally, to prevent model overfitting, the dropout layer of the model uses a relatively high threshold of 0.5.

[0065] In some embodiments, based on the root system feature image, root system phenotypic parameters are extracted, including:

[0066] Perform binarization processing on the root system probability distribution map to obtain a root system binary image;

[0067] Based on the root system binary image, count the number of root system pixel points;

[0068] Calculate the projected area according to the number of root system pixel points.

[0069] As shown in Figure 2 , the pixel value of each pixel point in the root system feature image represents the probability that the pixel point is a root system. Based on this, set a root system threshold, and perform binarization processing on the root system probability distribution map according to the root system threshold to obtain a root system binary image. For example, set the root system threshold to 5. Pixel points in the root system feature image with pixel values greater than or equal to 5 are binarized as root system pixel points to 255, and pixel points with pixel values less than 5 are binarized as soil pixel points to 0, obtaining a root system binary image. The white part represents the root system, and the black part represents the soil.

[0070] Based on the root system binary image, count the number of all root system pixel points, that is, count the number of pixel points with pixel value 255. According to the number of root system pixel points and the size of the root system image, calculate the projected area. The calculation method is to determine the area ratio according to the length, width, and total number of pixel points of the root system image, and determine the projected area according to the area ratio and the number of root system pixel points, which can be expressed as:

[0071]

[0072] Among them, S is the projected area, a is the length of the root system image, b is the width of the root system image, N is the total number of pixels in the root system image, and M is the number of root system pixel points.

[0073] In some embodiments, based on the root system feature image, the extraction of root system phenotypic parameters further includes:

[0074] Perform skeleton extraction processing on the root system binary image to obtain a root system skeleton image;

[0075] Based on the root system skeleton image, extract the root length, root diameter, and number of root tips.

[0076] In this embodiment, a skeleton extraction process is performed on the binary image of the root system to obtain a root system skeleton image. Based on the root system skeleton image, root system phenotypic parameters such as root length, root diameter, and number of root tips are further extracted.

[0077] Combined with Figure 3 As shown, the method for extracting the root length based on the root system skeleton image is:

[0078] According to the root system skeleton image, determine the center line of the root system;

[0079] Count the number of pixel points on the center line of the root system as the total root length.

[0080] In this embodiment, the center line of the root system in the binary image of the root system is extracted by using the skeleton extraction processing method ( Figure 3 each white line in it) as the center line of the root system, count the number of all pixel points on the center line of the root system, and take the counted number of pixel points as the total root length.

[0081] In some embodiments, after determining the center line of the root system, the root diameter can also be extracted. The method is;

[0082] Determine the position of the center line of the root system in the root system image. Taking each pixel point on the center line of the root system as the center of a circle, and the distance from the center of the circle to the boundary point between the root system and the soil as the radius to draw a circle, and take the diameter of the circle as the root diameter at the root system position corresponding to this pixel point.

[0083] In this embodiment, the thickness of the root system is different at different positions, and the corresponding root diameters are different. After determining the center line of the root system, based on the root system image, different pixel points on the center line of the root system represent different positions of the root system. For each pixel point, draw a circle with the pixel point as the center of the circle and the distance from the center of the circle to the boundary point between the root system and the soil as the radius, and take the diameter of the circle as the root diameter at the root system position corresponding to this pixel point. In this way, after drawing circles for all pixel points on the center line of the root system, the root diameters at different positions on the root system can be obtained.

[0084] In some embodiments, after determining the root diameters at different positions of the root system, the root diameters are classified according to the size of the root diameters. For example, the root diameters are divided into multiple levels such as 0 - 0.05 mm, 0.05 - 0.1 mm, 0.1 - 0.15 mm, 0.15 - 0.2 mm, etc. At the same time, root system phenotypic parameters such as the root length and average diameter of the root diameters at different levels can be counted.

[0085] In some embodiments, after determining the center line of the root system, the number of root tips can also be extracted. The method is;

[0086] Count the number of end points on the center line of the root system;

[0087] Determine the number of root tips according to the number of end points.

[0088] In this embodiment, based on the determined root centerline, the number of endpoints on the root centerline is counted, and the counted number of endpoints minus one is the number of root tips. In some ways, considering that there are irregular protrusions or depressions on the root surface, etc., to avoid misidentifying the irregular parts on the root surface as endpoints, resulting in deviations in the statistics of the number of root tips and root length, the root skeleton image is smoothed, and based on the smoothed root skeleton image, the number of endpoints on the root centerline is counted to determine the number of root tips.

[0089] In some embodiments, after extracting the root length, root diameter, and number of root tips, root phenotypic parameters such as root volume and root surface area are further calculated. Specifically, the average root diameter is calculated according to the root diameters at different positions of the root system; according to the total root length and the average root diameter, following the method of calculating the volume of a cylinder, with the total root length as the height and half of the average root diameter as the bottom radius, the total volume is calculated; according to the total root length and the average root diameter, following the method of calculating the surface area of a cylinder, the total surface area is calculated.

[0090] Furthermore, based on the classified root diameters, root phenotypic parameters such as the classified root length, surface area, projected area, average diameter, and volume are counted. For example, classified by 0.1 mm thickness, the root diameters of 0 - 0.1 mm are classified as the first level, the root diameters of 0.1 - 0.2 mm are classified as the second level, etc., and the root phenotypic parameters after classification such as the root length, projected area, average diameter, volume, surface area, volume, etc. of the first level are counted.

[0091] In some embodiments, after extracting the root phenotypic parameters, it further includes:

[0092] Calibrate the root phenotypic parameters according to a preset calibration relationship.

[0093] In this embodiment, considering that the effects of root system images at different resolutions are different, the boundary between the root system and the soil in a high - resolution root system image is clearer, finer roots can be identified, and the extracted root length is closer to the actual root length. However, the processing speed of high - resolution root system images will decrease; the boundary clarity between the root system and the soil in a low - resolution root system image is lower, and the extracted root length is shorter and the accuracy is lower. In the experiment, through multiple measurements and data statistics, a specific correspondence relationship can be established between the root phenotypic parameters of high - resolution root system images and those of low - resolution root system images. In this way, when it is necessary to improve the batch image processing speed and efficiency, a low - resolution root system image can be obtained. After processing the low - resolution root system image to obtain the root phenotypic parameters, the specific correspondence relationship is used as the calibration relationship between the high - resolution root system image and the low - resolution root system image, and the calibration relationship is used to calibrate the root phenotypic parameters, so as to obtain the root phenotypic parameters of the high - resolution root system image, improve the data processing efficiency, and ensure accuracy at the same time.

[0094] Such asFigure 4A As shown, in some embodiments, the root length calibration relationship between the low-resolution root system image (300 dpi) and the high-resolution root system image (600 dpi) is y = 1.1748x. The low-resolution root system image can be acquired, and after processing the low-resolution root system image, the root length is obtained as the parameter value of x. Through the root length calibration relationship, the root length y of the high-resolution root system image is obtained, thereby calibrating the root length. As Figure 4B As shown, the projection area calibration relationship between the low-resolution root system image and the high-resolution root system image is y = 0.7222x. The low-resolution root system image can be acquired, and after processing the low-resolution root system image, the projection area is obtained as the parameter value of x. Through the projection area calibration relationship, the projection area y of the high-resolution root system image is obtained, thereby calibrating the projection area. As Figure 4C As shown, the average diameter calibration relationship between the low-resolution root system image and the high-resolution root system image is y = 1.1708x - 0.3516. The low-resolution root system image can be acquired, and after processing the low-resolution root system image, the average diameter is obtained as the parameter value of x. Through the average diameter calibration relationship, the average diameter y of the high-resolution root system image is obtained, thereby calibrating the average diameter. As Figure 4D As shown, the root tip number calibration relationship between the low-resolution root system image and the high-resolution root system image is y = 2.1135x. The low-resolution root system image can be acquired, and after processing the low-resolution root system image, the root tip number is obtained as the parameter value of x. Through the root tip number calibration relationship, the root tip number y of the high-resolution root system image is obtained, thereby calibrating the root tip number.

[0095] In some embodiments, after obtaining the binary root system image, it further includes:

[0096] Filtering out interfering pixel points according to preset filtering conditions.

[0097] In this embodiment, considering the possible interfering factors in the soil environment and the irregularity of the root surface, in order to avoid the interference affecting the accuracy of the extracted parameters, some filtering conditions can be set, and the image is processed according to the filtering conditions during the image processing to filter out the interfering points. For example, there may be soil noise in the binary root system image, and there may be small protrusions on the root surface. Set the minimum projection area, and filter out the pixel points with a projection area smaller than the minimum projection area and do not regard them as roots for subsequent processing.

[0098] In some embodiments, after extracting various root phenotypic parameters, corresponding root phenotypic images can be drawn based on the various root phenotypic parameters. For example, after grading root diameters based on root diameters at different locations on the root system, a root diameter grading image can be drawn; after counting the number of root tips, a root tip distribution map can be drawn; after counting the number of endpoints on the root system centerline, the number of pixels between the two endpoints can be counted as the length of each root, and a horizon length image can be drawn, in which the length of each root is identified.

[0099] The effects of the root system image processing method according to the embodiments of this specification are described below with reference to specific examples.

[0100] A neural network model was trained using images of maize roots in a farmland and meadow vegetation roots to develop a root image processing model. During use, maize root images at different growth stages and time periods were acquired using scanning devices with resolutions of 300dpi and 600dpi, respectively. These images were then fed into the root image processing model to generate root characteristic images, from which various root phenotypic parameters were extracted.

[0101] like Figure 5A 、 5B As shown, the root length extracted by the method of this embodiment is compared with the root length identified by the naked eye of researchers. The method of this embodiment processes the root system image and extracts the root length in about 20 seconds, while the time it takes for researchers to identify the root length is about 20 hours. The root length extracted by the method of this embodiment is more accurate. Figures 6A - 6C As shown, the method of this embodiment is used to process the root system images collected continuously over a period of time, extract the root system phenotypic parameters through image processing, and calculate the changes of the parameters over a period of time. Figures 7A - 7D As shown, using the method of this embodiment to process root images of different resolutions can obtain different root phenotypic parameters. Comparative analysis shows that high-resolution root images can more accurately identify root length and a greater number of root tips, while low-resolution root images can obtain larger projected areas and root diameters, indicating that high-resolution root images provide more accurate identification results. However, considering data processing efficiency, low-resolution root images can be obtained and subsequently calibrated to obtain more accurate parameters. As shown in Tables 1 and 2, the root phenotypic parameters that can be obtained using the method of this embodiment include at least total root length, projected area, surface area, volume, number of root tips, root diameter, average diameter, root diameter after hierarchical statistics, and root length, surface area, projected area, and volume corresponding to different root diameter levels.

[0102] Table 1 Root phenotypic parameters

[0103]

[0104] Table 2 Root diameter classification

[0105]

[0106] In summary, the root system image processing method of this embodiment can quickly, efficiently, and accurately identify the root system and extract various root phenotype parameters, greatly improving the efficiency of scientific research work and providing a data basis for root system research.

[0107] It should be noted that the method of one or more embodiments of this specification can be executed by a single device, such as a computer or a server. The method of this embodiment can also be applied to a distributed scenario and completed by multiple devices cooperating with each other. In such a distributed scenario, one of the multiple devices can only execute one or more steps of the method of one or more embodiments of this specification, and these multiple devices will interact with each other to complete the described method.

[0108] It should be noted that the above describes specific embodiments of this specification. Other embodiments are within the scope of the appended claims. In some cases, the actions or steps recited in the claims can be executed in a different order than in the embodiments and still achieve the desired result. Additionally, the processes depicted in the figures do not necessarily require the specific order or sequential order shown to achieve the desired result. In certain embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0109] As Figure 8 shown, an embodiment of this specification also provides a root system image processing device, including:

[0110] An acquisition module, configured to acquire a root system image;

[0111] An identification module, configured to process the root system image using a root system image processing model to obtain a root system feature image;

[0112] An extraction module, configured to extract root phenotype parameters based on the root system feature image.

[0113] For the convenience of description, when describing the above device, it is divided into various modules according to functions and described separately. Of course, when implementing one or more embodiments of this specification, the functions of each module can be implemented in the same or multiple software and / or hardware.

[0114] The device of the above embodiment is used to implement the corresponding method in the foregoing embodiment and has the beneficial effects of the corresponding method embodiment, which will not be elaborated here.

[0115] Figure 9FIG. 0 shows a more specific schematic diagram of the hardware structure of the electronic device provided in this embodiment. The device may include: a processor 1010, a memory 1020, an input / output interface 1030, a communication interface 1040, and a bus 1050. Among them, the processor 1010, the memory 1020, the input / output interface 1030, and the communication interface 1040 are communicatively connected to each other inside the device through the bus 1050.

[0116] The processor 1010 may be implemented in a general-purpose CPU (Central Processing Unit), a microprocessor, an application-specific integrated circuit (ASIC), or one or more integrated circuits, etc., and is used to execute relevant programs to implement the technical solutions provided in the embodiments of this specification.

[0117] The memory 1020 may be implemented in the form of a ROM (Read Only Memory), a RAM (Random Access Memory), a static storage device, a dynamic storage device, etc. The memory 1020 may store an operating system and other application programs. When implementing the technical solutions provided in the embodiments of this specification through software or firmware, the relevant program codes are stored in the memory 1020 and are called and executed by the processor 1010.

[0118] The input / output interface 1030 is used to connect to an input / output module to implement information input and output. The input / output module may be configured as a component in the device (not shown in the figure) or externally connected to the device to provide corresponding functions. Among them, the input device may include a keyboard, a mouse, a touch screen, a microphone, various sensors, etc., and the output device may include a display, a speaker, a vibrator, an indicator light, etc.

[0119] The communication interface 1040 is used to connect to a communication module (not shown in the figure) to implement communication interaction between this device and other devices. Among them, the communication module may implement communication in a wired manner (such as USB, network cable, etc.) or in a wireless manner (such as mobile network, WIFI, Bluetooth, etc.).

[0120] The bus 1050 includes a path for transmitting information between various components of the device (such as the processor 1010, the memory 1020, the input / output interface 1030, and the communication interface 1040).

[0121] It should be noted that although the above device only shows the processor 1010, the memory 1020, the input / output interface 1030, the communication interface 1040, and the bus 1050, in the specific implementation process, the device may also include other components necessary for normal operation. In addition, those skilled in the art can understand that the above device may also only include the components necessary to implement the solution of the embodiments of this specification, and does not necessarily include all the components shown in the figure.

[0122] The electronic device in the above embodiment is used to implement the corresponding method in the foregoing embodiment, and has the beneficial effects of the corresponding method embodiment, which will not be elaborated here.

[0123] The computer-readable medium of this embodiment includes permanent and non-permanent, removable and non-removable media, and information storage can be implemented by any method or technology. The information can be computer-readable instructions, data structures, program modules, or other data. Examples of computer storage media include, but are not limited to, phase change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technologies, compact disc read-only memory (CD-ROM), digital versatile disc (DVD) or other optical storage, magnetic cassette tapes, magnetic tape magnetic disk storage or other magnetic storage devices, or any other non-transmission medium that can be used to store information that can be accessed by a computing device.

[0124] Those of ordinary skill in the art should understand that: the discussion of any of the above embodiments is only exemplary, and is not intended to imply that the scope of the present disclosure (including the claims) is limited to these examples; under the concept of the present disclosure, the technical features in the above embodiments or different embodiments can also be combined, the steps can be implemented in any order, and there are many other variations in different aspects of one or more embodiments of the present specification as described above, and they are not provided in detail for the sake of brevity.

[0125] In addition, for simplicity of explanation and discussion, and so as not to make one or more embodiments of this specification difficult to understand, well-known power / ground connections to integrated circuit (IC) chips and other components may or may not be shown in the provided drawings. Further, the devices may be shown in block diagram form in order to avoid making one or more embodiments of this specification difficult to understand, and this also takes into account the fact that details regarding the implementation of these block diagram devices are highly dependent on the platform on which one or more embodiments of this specification are to be implemented (i.e., these details should be entirely within the understanding of those of ordinary skill in the art). In cases where specific details (such as circuits) are set forth to describe exemplary embodiments of the present disclosure, it will be apparent to those of ordinary skill in the art that one or more embodiments of this specification may be practiced without these specific details or with variations of these specific details. Accordingly, these descriptions should be regarded as illustrative rather than restrictive.

[0126] Although the present disclosure has been described in connection with specific embodiments thereof, many alternatives, modifications, and variations of these embodiments will be apparent to those of ordinary skill in the art based on the foregoing description. For example, other memory architectures (such as dynamic RAM (DRAM)) may be used with the embodiments discussed.

[0127] One or more embodiments of this specification are intended to cover all such alternatives, modifications, and variations that fall within the broad scope of the appended claims. Accordingly, any omissions, modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of one or more embodiments of this specification shall be included within the scope of protection of the present disclosure.

Claims

1. A root system image processing method, characterized in that, Including: Obtaining root system images; Processing the root system images using a root system image processing model to obtain root system feature images; wherein, the root system feature images are root system probability distribution maps; Based on the root system feature images, extracting root system phenotypic parameters, including: performing binarization processing on the root system probability distribution map to obtain a root system binary image; based on the root system binary image, counting the number of root system pixel points; according to the number of root system pixel points and the size of the root system image, calculating the projected area; performing skeleton extraction processing on the root system binary image to obtain a root system skeleton image; based on the root system skeleton image, extracting the root length, root diameter, and number of root tips, and determining the root system center line; counting the number of pixel points on the root system center line as the total root length; according to the total root length and root diameter, calculating the total volume and total surface area; wherein, the method for calculating the projected area is: Wherein, S is the projected area, a is the length of the root system image, b is the width of the root system image, N is the total number of pixels in the root system image, and M is the number of root system pixel points; Calibrating the root system phenotypic parameters according to a preset resolution correspondence relationship, including: Processing the low-resolution root system image obtained to obtain the root length, projected area, average diameter, and number of root tips; Determining the root length of the high-resolution root system image according to a preset root length calibration relationship; wherein, the root length calibration relationship is: y = 1.1748x, x is the root length corresponding to the low-resolution root system image, and y is the root length of the calibrated high-resolution root system image; Determining the projected area of the high-resolution root system image according to a preset projected area calibration relationship; wherein, the projected area calibration relationship is y = 0.7222x, x is the projected area corresponding to the low-resolution root system image, and y is the projected area of the calibrated high-resolution root system image; Determining the average diameter of the high-resolution root system image according to a preset average diameter calibration relationship; wherein, the average diameter calibration relationship is y = 1.1708x - 0.3516, x is the average diameter corresponding to the low-resolution root system image, and y is the average diameter of the calibrated high-resolution root system image; Determining the number of root tips of the high-resolution root system image according to a preset number of root tips calibration relationship; wherein, the number of root tips calibration relationship is y = 2.1135x, x is the number of root tips corresponding to the low-resolution root system image, and y is the number of root tips of the calibrated high-resolution root system image.

2. The method according to claim 1, wherein After determining the root system center line, further including: Determining the position of the root system center line in the root system image, taking each pixel point on the root system center line as the center of a circle, and using the distance from the center of the circle to the boundary point between the root system and the soil as the radius to draw a circle, and taking the diameter of the circle as the root diameter at the position corresponding to the pixel point.

3. The method according to claim 2, characterized in that After determining the root diameter, further including: Dividing the root diameter grades according to the root diameters of different pixel points on the root system center line.

4. The method according to claim 1, wherein After determining the root system center line, further including; Counting the number of end points on the root system center line; Determining the number of root tips according to the number of end points.

5. The method according to claim 1, characterized in that, After extracting the root length, root diameter, and number of root tips, further including: Calculating the root volume and root surface area according to the root length and root diameter.

6. The method according to claim 1, characterized in that After obtaining the binary image of the root system, it further includes: Filtering out interfering pixel points according to a preset filtering condition.

Images